Skip to main content
newspals
Topics
Concepts
Editors
Newsletter
English
Mixture Of Experts — Concepts | NewsPals
Concepts
·
Mixture Of Experts
the lore behind the feed
Mixture Of Experts
The stories that keep pulling this idea back into the feed.
4 stories
In the feed
ai-ml
Nvidia's Nemotron 3.5 Lightning Shifts Open Weights From Size Flex to Deployability
A 30B MoE model with 3B active parameters, 1M context, OpenMDW-1.1 licensing, and recipes makes deployment the story.
ai-ml
Ling 3.0 Flash makes 124B parameters about active compute speed, not size
Ant Group and InclusionAI are selling an open weights MoE around 5.1B active parameters per token, which is the point.
ai-ml
Inkling Small Hits 40, Nearly Matching Inkling 41 With 276B Parameters
Thinking Machines Lab’s smaller open weights model uses 276B total and 12B active MoE, while Inkling uses 975B total and 41B active.
Also vibing
Agentic AI
Ant Group
Artificial Analysis
InclusionAI
Inkling Small
Ling 3.0 Flash
Nvidia Nemotron
NVIDIA NIM